RCy3: adding and removing table columns within a for loop

[Márton Ölbei]

Hello everyone,

I would like to ask a question about something I encountered using RCy3.

I have a network I would like to visualize under multiple circumstances, using multiple (20+) datasets.

I have written an RCy3 script which reads in the base network and visualizes it. I have set up a loop which iterates over my data files, and adds them to the network with the loadTableData command. Essentially, these are just gene identifiers with a p.value.

What my script does/I would like it to do:

- read in dataset file

- highlight the IDs in the networks, select their first neighbours

- create a new subnetwork out of this selection

- visualize the p.value using a continuous mapping for size and colour

- save a snapshot

- revert to the base network

- delete added columns

My problem is: when I add/delete columns something happens to the p.values, I get back values and IDs that aren't even in my original input files.

Here's an example. This is one of my input files:

Node    name    p    p-corr
SL1344_2856    None    6.3101e-6    3.7861e-5
SL1344_3325    None    5.9315e-6    3.7861e-5
SL1344_3323    None    5.9315e-6    3.7861e-5
SL1344_0069    None    4.3396e-6    3.7861e-5
SL1344_0057    None    1.0797e-3    3.2392e-3
SL1344_0055    None    1.0797e-3    3.2392e-3
SL1344_1325    None    9.0257e-4    3.2392e-3
SL1344_0056    None    7.2570e-4    3.2392e-3
SL1344_0071    None    1.9829e-3    5.2878e-3
SL1344_0075    None    2.5288e-3    6.0690e-3
SL1344_0072    None    4.5240e-3    9.8705e-3
SL1344_1591    None    5.8982e-3    1.1796e-2
SL1344_4525    None    6.7170e-3    1.2401e-2
SL1344_3358    None    2.8175e-2    4.8300e-2
SL1344_0074    None    4.2529e-2    6.8047e-2
SL1344_0076    None    5.0603e-2    7.1440e-2
SL1344_0070    None    5.0603e-2    7.1440e-2
SL1344_3433    None    5.8645e-2    7.8193e-2
SL1344_0077    None    6.6333e-2    7.9600e-2
SL1344_0068    None    6.6333e-2    7.9600e-2
SL1344_0067    None    7.3993e-2    8.4563e-2
SL1344_1169    None    2.8198e-1    2.9424e-1
SL1344_0079    None    2.7181e-1    2.9424e-1

What I get as the output in Cytoscape for the same file, after loading the table through RCy3 with the loadTableData command:

shared name    p    p.corr
SL1344_1695    0.35802    0.46542
SL1344_1169    0.28198    0.37598
SL1344_4525    0.14795    0.20245
SL1344_0610    0.1103    0.15502
SL1344_1747    0.1103    0.15502
SL1344_0608    0.088909    0.13598
SL1344_0611    0.081517    0.12845
SL1344_0067    0.073993    0.12412
SL1344_0068    0.066333    0.12412
SL1344_0182    0.073993    0.12412
SL1344_0183    0.066333    0.12412
SL1344_3875    0.037434    0.12412
SL1344_3081    0.073993    0.12412
SL1344_3302    0.042529    0.12412
SL1344_2856    0.07371    0.12412
SL1344_3303    0.042529    0.12412
SL1344_1591    0.027056    0.12412
SL1344_4008    0.069442    0.12412
SL1344_3358    0.0062351    0.040528
SL1344_0675    0.0022262    0.019293
SL1344_3433    0.0014047    0.018262
SL1344_3469    5.72E-04    0.014864
SL1344_3332    1.54E-04    0.0080026

Not only are they different IDs, but they have completely different p.corr values as well. In fact, if I grep my input file directory for any of these values I get no hits back.

If I import the table manually I get the correct values back.

Here is my script:

library(RCy3)
library(RColorBrewer)
# -------- Setting up --------

setwd("/Users/olbeim/Documents/ContextTool_analysis/2_pipeline/visualization/")
salmonet <- read.csv("test_network.csv", header = TRUE, sep=',', stringsAsFactors = FALSE)
nodelist <- unique(c(salmonet$Node_1_locus_tag, salmonet$Node_2_locus_tag))
nameTable <- read.csv("/Users/olbeim/Documents/ContextTool_analysis/2_pipeline/lt_to_genename", sep='\t', header=TRUE, stringsAsFactors = FALSE)
nodes <- data.frame(id=nodelist, # integers #load data
                    stringsAsFactors=FALSE)
edges <- data.frame(source=salmonet$Node_1_locus_tag,
                    target=salmonet$Node_2_locus_tag,
                    layer = salmonet$Layer,# numeric
                    stringsAsFactors=FALSE)

createNetworkFromDataFrames(nodes,edges, title="SalCom network", collection="SalCom") #load network
loadTableData(nameTable, data.key.column = "LT", table.key.column = "name")#add gene names
setNodeColorDefault('#C0C0C0') # set to gray
setNodeLabelMapping(table.column = 'Gene') #set to gene name
fitContent() #make sure everyting fits

# -------- CHAT output pcorr size -------
files = list.files("~/Documents/ContextTool_analysis/3_output/CHAT_output/nodes25/", pattern="salcom_*", recursive=FALSE, full.names=TRUE) #read data files

for (i in files) {
  nameTable <- read.csv(file = i, sep='\t', header=TRUE, stringsAsFactors = FALSE)
  loadTableData(nameTable, data.key.column = "Node", table.key.column = "shared name") #add data table to network
  selectNodes(nameTable$Node, by.col = "id", preserve.current.selection = FALSE)
  setNodeSizeMapping('p.corr', c(min(nameTable$p.corr), max(nameTable$p.corr)), c(75, 20))
  setNodeColorMapping('p.corr', c(min(nameTable$p.corr), max(nameTable$p.corr)), c('#5577FF','#FFFFFF'))
  selectFirstNeighbors(direction="any")
  createSubnetwork(nodes="selected", subnetwork.name = i)
  clearSelection()

  fitContent()

  exportImage(filename = paste(i,'_analysis.JPEG', sep=""), resolution = 72, zoom = 500)
  setCurrentNetwork(52) #revert to original network using its SUID
  deleteTableColumn(column = "Node")
  deleteTableColumn(column = "p")
  deleteTableColumn(column = "p.corr") #delete columns before reading in the next data file
}

I assume something goes wrong when I'm adding the tables. I realize this has been a long post, I am just trying to be as transparent with my problem as I possibly can.

I would really appreciate if someone could help me out with this. I am really struggling to understand why those IDs and pvalues show up when a., they shouldn't b., some of them don't exist.

I wonder, is it possible that it has something to do with the kind of information (e.g. string, float etc.) the columns are set to by default?

Best wishes,

Marton

[Alex Pico]

Hi Márton,

Thanks for the detailed problem description. It really helps!

I have created some test files based on your descriptions and examples below. However, these seem to work just fine…. So, I think I might still be missing something particular to your files and setup, or I’m not fully understanding what you are attempting.

Try running my files (attached) and see if they work as you would expect.

* If yes, then see if you can spot the differences to adapt your script.

* If no, then please send sample files I can work with.

Things I adapted:

1. I’m using csv instead of tsv (shouldn’t make a difference, unless error is in R)

2. Creating network from edges alone. This is a nice short cut if all of your nodes have at least one connection. If you have unconnected nodes, then you’ll need to specify them explicitly. (Again, not critical to your problem; just a tip)

3. Collecting parent network SUID as time of creation (see parent.net variable), which can be used later in for-loop. (Just a tip)

4. Added lockNodeDimensions() and setNodeShapeDefault('ELLIPSE’) to the script to make the size control work.

5. Changed column in “salcom” files from “Name” to “Label”, otherwise it will overwrite your node names irreversibly. See Node Table.

6. Added layoutNetwork() in for-loop, just as a suggestion

7. Removed unnecessary steps.

Let me know if any of this doesn’t make sense.

 - Alex

[Márton Ölbei]

Hi Alex,

Thank you very much for your help, your script is working marvelously! I haven't yet figured out what broke mine, but I will get back to you once I figure it out.

Thank you very much for the help again!

Best wishes,

Marton

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/wCyUSp5uW0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/2C4EC588-ABA6-410C-A8FF-45DB21696240%40gladstone.ucsf.edu.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/wCyUSp5uW0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/2C4EC588-ABA6-410C-A8FF-45DB21696240%40gladstone.ucsf.edu.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/wCyUSp5uW0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/2C4EC588-ABA6-410C-A8FF-45DB21696240%40gladstone.ucsf.edu.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/wCyUSp5uW0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/2C4EC588-ABA6-410C-A8FF-45DB21696240%40gladstone.ucsf.edu.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/wCyUSp5uW0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/2C4EC588-ABA6-410C-A8FF-45DB21696240%40gladstone.ucsf.edu.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/wCyUSp5uW0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/2C4EC588-ABA6-410C-A8FF-45DB21696240%40gladstone.ucsf.edu.
For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/dd79fd8e-bd2a-4632-9dbe-1e633ce6bd0f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "cytoscape-helpdesk" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cytoscape-helpdesk/wCyUSp5uW0I/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cytoscape-helpd...@googlegroups.com.
To post to this group, send email to cytoscape...@googlegroups.com.
Visit this group at https://groups.google.com/group/cytoscape-helpdesk.
To view this discussion on the web visit https://groups.google.com/d/msgid/cytoscape-helpdesk/2C4EC588-ABA6-410C-A8FF-45DB21696240%40gladstone.ucsf.edu.
For more options, visit https://groups.google.com/d/optout.